建设工程造价指数预测研究

Research on Prediction of Construction Cost Index

建设工程造价指数是反映工程价格波动的重要指标,是进行工程造价管理的重要依据。针对灰色预测模型和人工神经网络的局限性,在引入适用性更强的随机振荡GM(1,1)模型和蚁群改进BP神经网络的基础上,提出将两者融合的组合循环模型并进行实证分析验证,为准确预测工程造价指数提供借鉴。

The construction cost index is an important indicator reflecting the construction cost fluctuations,and is an important basis for effective management of the construction cost. To overcome the limitations of gray prediction model and artificial neural networks,on the basis of introducing more applicable stochastic oscillation GM(1,1)model and BP neural network improved by ant colony algorithm,the paper proposes the combined cycle model of them,validate the model through an example analysis,and provides a new reference for the prediction of construction cost index.